Teacher Talk in Middle School History: Differences Across Instructional Conditions and Their Impact on Seventh Graders’ Learning

Abstract

Middle school social studies classrooms present unique challenges and opportunities for promoting disciplinary understanding and school-relevant language for all students in linguistically diverse settings. Discussion-based pedagogies have been shown to support student learning in these contexts, but, besides being underused, the hypothesis that teacher talk might mediate this relation remains untested. Thus, this dissertation developed and applied the History Teacher Talk Coding System (HTT), a novel instrument for analyzing the quantity and quality of cross-disciplinary and history-specific teacher talk through three key levers of responsive instruction for linguistically diverse classrooms: general discussion, historical reasoning, and language development. The data for this dissertation’s three studies are drawn from two randomized control trials (RCTs) testing the World Generation (WorldGen) discussion-based social studies curriculum in classrooms serving large proportions of Emergent Bilinguals (Capin et al., under preparation; Martinez et al., 2024). The WorldGen studies featured a within-teacher design, in which teachers’ seventh-grade classes were randomly assigned to either treatment or comparison (i.e., teachers’ typical) instruction. Study 1 developed the HTT and examined its sensitivity for detecting variability in teacher talk. Line-by-line coding of teachers’ instruction across instructional conditions indicated that the HTT was sensitive to differences in teacher talk by condition, teacher, and lesson. Teachers facilitated greater and richer use of General Discussion and Historical Reasoning moves during treatment lessons; however, little variation was noted in Vocabulary Instruction across conditions. In Study 2, I applied the HTT to a larger sample. Findings revealed that—relative to comparison instruction—when teachers used WorldGen, they and their students engaged in more Instructional Talk, and teachers fostered greater discussion, provided additional scaffolding for historical reasoning, and increased student-centered vocabulary instruction. Finally, in Study 3, I used hierarchical linear modeling to replicate the WorldGen intervention’s main effects and test whether the quantity and quality (both frequency and diversity) of teacher talk mediated the intervention’s effects on students’ content knowledge, vocabulary, and disciplinary literacy (i.e., historical thinking). Treatment effects replicated those in the larger RCT, and the quantity and, particularly, the quality of teacher talk mediated the curriculum’s impact on students’ content and language outcomes. Altogether, this dissertation offers a novel methodological tool that has generated empirical evidence for advancing the field’s understanding of discussion-rich, linguistically responsive teaching in social studies. These findings underscore the mediating effect of high-quality, multidimensional teacher talk on the simultaneous development of students’ content knowledge and language proficiency.